Introduction

Rape is a significant problem on college campuses in the United States where it is escalated by norms of heavy drinking (Cole 2006; Locke and Mahalik 2005; Mohler-Kuo et al. 2004). Although many college students will be affected by rape as either a victim themselves or a friend or family member of a victim, only some students will engage in efforts to end rape and/or help survivors of rape. Who participates in anti-rape collective action on college campuses? Who helps survivors of rape?

Past research on sexual violence suggests that college students’ reactions to rape are complex, involving attitudes (Burt 1980; Liss et al. 2004; Payne et al. 1999; Peterson and Muehlenhard 2004), attributions of fault (Barnett et al. 1992; Brems and Wagner 1994; Ford et al. 1998; Grubb and Harrower 2008; Kanekar and Seksaria 1993; Krulewitz and Nash 1979; Pollard 1992), and emotions (for a review, see Ahrens and Campbell 2000). Additionally, past research highlights gender differences in college students’ attitudes (Anderson et al. 1997; Twenge 1997), attributions of fault (Gillen and Muncer 1995), and emotions (Brownmiller 1975) regarding rape. Although attitudes, attributions of fault, and emotions have been studied individually as important factors that are related to college students’ behavioral responses to rape, they have not been examined simultaneously. Given that these psychological factors co-occur in the real world, research that simultaneously examines their ability to predict behavioral responses to rape may be useful for researchers who aim to develop programs to increase behaviors that will reduce the frequency of rape and minimize its deleterious consequences for survivors. Additionally, research that explores whether the processes whereby these factors relate to behavioral responses to rape differ by gender may further help researchers tailor their programs for female and male college students.

In the current research, we sought to address these gaps by asking American college students to read about an incident of rape and measuring their responses to this incident. We then examined mean differences in attitudes, attributions of fault, and emotions between female and male college students using a multivariate analysis of variance. We further examined the degree to which these factors predicted their desire to engage in anti-rape collective action and reported likelihood of helping a survivor of rape as well as whether gender moderated these relationships with two regression analyses. Before describing the current study, we review relevant research and theory related to attitudes, attributions of fault, and emotions. Unless otherwise noted, the research reviewed was conducted with samples of U.S. college students.

Attitudes: Feminism and Rape Myth Acceptance

Attitudes towards feminism and rape myth acceptance encompass two trait-level constructs that may be related to college students’ responses to rape. The link between feminist attitudes and engagement in feminist collective action is well established among samples of U.S. college students (Liss et al. 2004; Nelson et al. 2008) and U.S. women (Duncan 1999). People who hold feminist attitudes may be more likely to engage in anti-rape collective action because they emphasize societal-level contributors to sexual violence. Feminists have highlighted the connection between power inequalities that exist between men and women in society and the widespread acceptance of violence against women (Cowan and Quinton 1997). They have identified representations that reduce women to sex objects as both creating and supporting attitudes accepting of violence and discrimination toward women (MacKinnon 1986). Further, feminists have examined the ways in which society’s portrayal of heterosexuality provides a “cultural scaffolding of rape” by normalizing unwanted and even coercive sex (Gavey 2005). Importantly, there is a relationship between attitudes towards feminism and gender: Women tend to hold stronger feminist attitudes than men (Twenge 1997). Therefore, in the current work we expected that female college students would hold stronger feminist attitudes than male college students. We further expected that college students’ feminist attitudes would be a predictor of their desire to engage in anti-rape collective action, but not reported likelihood of helping a survivor of rape, because they emphasize the societal roots of sexual violence.

Rape myth acceptance attitudes may also influence responses to rape. Burt (1980) defined rape myths as “prejudicial, stereotyped, or false beliefs about rape, rape victims, and rapists” which ultimately create a “climate hostile to rape victims” (p. 217). Rape myths encourage responsibility for rape to be placed on survivors of assault by focusing on what survivors did to provoke the rape. Therefore, rape myths decrease perceptions that survivors deserve help by blaming them for the rape. Decreased perceptions of deservingness in turn diminish willingness to help others (Otten et al. 1988), suggesting that rape myths may diminish willingness to help rape survivors. Furthermore, people who hold strong rape myth acceptance attitudes may blame survivors for rape to the exclusion of larger, societal forces. Rape myth acceptance attitudes are also related to gender: Men endorse rape myth acceptance attitudes more than women (Anderson et al. 1997). We therefore expected that male college students would hold stronger rape myth acceptance attitudes than female college students. Because rape myths focus on the responsibility of individuals for rape rather than societal forces, we further expected that college students’ rape myth acceptance attitudes would predict their reported likelihood of helping a survivor of rape but not desire to engage in anti-rape collective action.

Attributions of Fault to Society, Perpetrator, and/or Survivor

Attributions of fault or blame for rape represent an additional variable that is likely related to how college students respond to rape. Unlike attitudes, which are trait-level constructs expected to be stable across situations, attributions of fault are state-level constructs which may vary across different situations. These attributions have been made at both societal and individual levels by college students in the U.S. (Cowan and Quinton 1997) and U.K. (Gillen and Muncer 1995). At the societal level, college students identify societal forces that influence the behavior of individuals involved in rape (Gillen and Muncer 1995). They draw connections between the attitude of society that date rape is unimportant with the attitude of males that date rape is unimportant, and expected norms of dates with misunderstandings by men of women’s behavior. Past work has shown that female college students in the U.K. are more likely to draw connections between societal causes and individual behaviors than male students (Gillen and Muncer 1995). We therefore expected that female college students in the current study would also attribute more fault to society than male college students. Further, people who identify society as a cause of rape likely acknowledge that rape will continue until societal changes occur. Therefore, we expected that college students’ attributions of fault to society would predict desire to engage in anti-rape collective action. We did not expect that their attributions of fault to society would predict reported likelihood of helping a survivor of rape.

In addition to blaming society for rape, many people attribute fault to the perpetrator of rape. U.K. college students in Gillen and Muncer’s (1995) study identified male drunkenness, the need by males for dominance, misunderstandings by men of women’s behavior, and males’ attitudes that date rape is unimportant as causes of date rape. Female college students in this study were more likely to endorse more of these reasons as causes of rape. Male college students, in contrast, were more likely to endorse only misunderstandings by men of women’s behavior as a cause of rape. Other work has also shown that female college students blame male perpetrators more than male college students blame male perpetrators (Krulewitz and Nash 1979). We therefore expected that female college students in the current study would attribute more fault to the male perpetrator than male college students. Attributing fault to the individual male perpetrator of rape may result in behavior at the individual level rather than the collective level. Therefore, we expected that college students’ attributions of fault to the male perpetrator would be a predictor of reported likelihood of helping a survivor of rape but not desire to engage in anti-rape collective action.

Finally, people often attribute fault to the female survivor. Female survivors are blamed for rape if they had been drinking alcohol at the time of the assault (Cowan and Quinton 1997; Gillen and Muncer 1995), acted flirtatious (Cowan and Quinton 1997; Gillen and Muncer 1995), wore revealing clothing (Cowan and Quinton 1997), had previous sexual partners (Kanekar and Seksaria 1993), did not physically resist their assailant (Kanekar and Seksaria 1993), and/or were raped by an acquaintance (Barnett et al. 1992; Grubb and Harrower 2008; Kanekar and Seksaria 1993). People who attribute fault to the female survivor believe that the woman did something wrong to provoke the rape. This belief exonerates both the male perpetrator and society as responsible for the rape, and undermines perceptions that the survivor deserves help (Otten et al. 1988). Additionally, past work has consistently found that men attribute more fault to female rape survivors than women (Ahrens and Campbell 2000; Barnett et al. 1992; Cowan and Quinton 1997; Grubb and Harrower 2008; Kanekar and Seksaria 1993). We therefore expected that male college students in the current study would attribute greater fault to the female survivor than female college students. Because it is an individual based attribution of fault, we also expected that college students’ attributions of fault to a survivor would predict reported likelihood of helping but not desire to engage in anti-rape collective action.

Emotions: Fear, Anger, and Pity

Emotions may also influence American college students’ responses to rape. Feelings of fear are a common response to increased awareness of rape (Ahrens and Campbell 2000; Hughes et al. 2003; Klaw et al. 2005). Fear is felt as a result of perceiving that the physical safety of oneself and/or one’s close others has been endangered (Cottrell and Neuberg 2005). In the current context, people may feel fear as the result of perceiving that they are at risk of experiencing rape and/or that their friends and family members are at risk. Fear motivates action to protect the self and close others from threats to physical safety, including rape (Cottrell and Neuberg 2005). For example, fear appeals focused on rape are related to women’s intentions to engage in self-protective strategies against rape as well as to men’s intentions to engage in strategies protective of female close others (Morrison 2005). Young women in the U.S. (Cobbina et al. 2008) and adult women in Europe and North America (Yodanis 2004) perceive themselves to be at greater risk of rape than men, and experience greater fear in response to awareness of rape than men. This research supports the theoretical claims of Susan Brownmiller (1975) and other feminist theorists (e.g., Fredrickson and Roberts 1997; Griffin 1979) who have emphasized women’s experience of fear in response to the threat of rape. We therefore expected that female college students would experience greater fear than male college students in the current study. Further, anti-rape collective action offers an opportunity to encourage societal changes to reduce the chance that the self and close others will experience sexual violence. Consequently, we also expected that fear would predict desire to engage in anti-rape collective action but not reported likelihood of helping a survivor of rape.

Feelings of anger have also been highlighted as a response to rape among U.S. college students (Ahrens and Campbell 2000) and U.S. rape victim advocates (Wasco and Campbell 2002). Anger is felt as a result of perceiving a threat to one’s personal freedoms and rights and/or those of close others (Cottrell and Neuberg 2005). Rape represents a threat to people’s rights to physical safety and control. In the current context, people may feel anger as the result of perceiving that their own rights and/or the rights of close others to physical safety and control have been threatened by the risk of rape. Feelings of anger have been highlighted as a motivator of collective action among U.S. adults (Smith et al. 2008) and Dutch college students (van Zomeren et al. 2004; van Zomeren and Lodewijkx 2005). Anger has also been linked to interpersonal helping behavior directed towards victims of crime (van Zomeren and Lodewijkx 2005). Women are aware of their increased vulnerability to rape (Griffin 1979), and may therefore be more likely than men to perceive that their rights to physical safety and control are threatened by rape. Consequently, we expected that female college students would experience greater anger than male college students in response to reading a scenario about rape in the current study. Further, collective action represents a way to protest threats to personal freedoms and rights, and helping represents a way to enhance the well-being of close others whose freedoms and rights have been violated. Therefore, we expected that anger would predict desire to engage in anti-rape collective action and reported likelihood of helping a survivor of rape.

Feelings of pity may also be felt in response to rape. Although they have not been a focus of sexual violence research, feelings of pity have been studied in relation to responses to victims of other types of crime (van Zomeren and Lodewijkx 2005). Pity is felt as a result of perceiving that a close other’s well-being has been compromised for reasons for which they are not responsible (Cottrell and Neuberg 2005). Rape represents a serious threat to the psychological and physical well-being of survivors (Elklit et al. 2009; Koss et al. 2002; McMullin and White 2006). Because we previously hypothesized that female college students would attribute less fault, or responsibility, to the female survivor, we also expected that female college students would experience more pity than male college students in the current study. Furthermore, pity felt as a result of perceiving that a close other’s well-being has been compromised motivates action to restore the close other’s well-being (Cottrell and Neuberg 2005). We therefore expected that pity would predict reported likelihood of helping a survivor of rape but not desire to engage in anti-rape collective action.

Current Study

In the current study, we explored how attitudes, attributions of fault, and emotions predict American college students’ desire to engage in anti-rape collective action and reported likelihood of helping a survivor of rape. We conducted a study in which American college students read a short description of a female college student’s experience of being raped and then answered a series of questions about their perceptions of and reactions to the incident described, as well as how they would respond to the rape. We next examined three sets of hypotheses. The first set of hypotheses is related to gender differences and includes:

  • Hypothesis 1a: Female college students will hold stronger attitudes towards feminism and weaker rape myth acceptance attitudes than male college students.

  • Hypothesis 1b: Female college students will attribute more fault to society, more fault to the male perpetrator, and less fault to the female survivor than male college students.

  • Hypothesis 1c: Female college students will experience greater fear, anger, and pity than male college students.

The second set of hypotheses is related to predictors of desire to engage in anti-rape collective action and includes:

  • Hypothesis 2a: Attitudes towards feminism positively predict desire to engage in anti-rape collective action.

  • Hypothesis 2b: Attributions of fault to society positively predict desire to engage in anti-rape collective action.

  • Hypothesis 2c: Feelings of fear and anger positively predict desire to engage in anti-rape collective action.

The third set of hypotheses is related to predictors of reported likelihood of helping a survivor of rape and includes:

  • Hypothesis 3a: Rape myth acceptance attitudes negatively predict reported likelihood of helping a survivor of rape.

  • Hypothesis 3b: Attributions of fault to a female survivor of rape negatively predict and attributions of fault to a male perpetrator positively predict reported likelihood of helping a survivor of rape.

  • Hypothesis 3c: Feelings of anger and pity positively predict reported likelihood of helping a survivor of rape.

Our hypotheses are based on prior research that has primarily examined these three factors individually. Therefore, we expected that some factors may play a stronger or weaker role in responses to rape when the remaining factors were accounted for. To evaluate the first set of hypotheses, we examined the descriptive statistics and correlations separately for female and male participants. To evaluate the second and third sets of hypotheses, we conducted regression analyses including the main effects of gender, attitudes, attributions of fault, and emotions in the first steps to demonstrate their ability to predict responses to rape while controlling for each other. The interaction of gender with each of the predictors was included in the second steps to test the role of gender as a moderator of these effects.

Although we hypothesized that female and male college students would score differently on the predictor variables, we did not hypothesize that gender would moderate the effects of the predictor variables. There is a substantial amount of theory and research to support our hypotheses that female and male college students would differ in the extent to which they endorse the predictor variables. Much of these gender differences are related to women’s awareness of their increased vulnerability to rape (Brownmiller 1975; Fredrickson and Roberts 1997; Griffin 1979). That is, given their substantial increased risk of experiencing rape, women’s rape-related attitudes, attributions of fault, and emotions should differ from those of men. In contrast, there is no known theory or research to suggest that the process linking these predictor variables with desire to engage in anti-rape collective action or reported likelihood of helping a survivor of rape would be differ according to gender. We therefore expected that the relationships between the predictor variables and dependent variables would be similar for male and female college students.

Method

Participants

One hundred and seventy-nine undergraduates from the University of Connecticut, a large public university in the northeast United States, participated in the study in the spring of 2007. Students enrolled in introductory psychology courses were eligible to participate for partial course credit. One hundred and five participants (58.7%) identified as female. Female participants were between the ages of 18 and 23 (M = 18.98, SD = .93). Female participants identified as White or European-American (76.2%), Asian or Asian-American (10.5%), Black or African-American (6.7%), Latino/a or Hispanic-American (4.8%), and “Other” (1.0%). The majority of the female participants (99%) identified as heterosexual and one identified as bisexual. Seventy-four participants (41.3%) identified as male. Male participants were between the ages of 18 and 23 (M = 19.18, SD = 1.23). Male participants identified as White or European-American (83.8%), Asian or Asian-American (8.1%), Latino/a or Hispanic-American (2.7%), Multiracial (2.7%), and “Other” (2.7%). The majority of male participants (98.6%) identified as heterosexual and one identified as “Other.”

Procedure

All students who completed a mass pre-screening at the beginning of the semester were eligible for the study. Included in the pre-screening were measures of rape myth acceptance attitudes and attitudes towards feminism. These variables were measured in advance because they represent trait-level phenomena that should not be influenced by the scenario described in the study. Additionally, by collecting these data separately we were able to ensure that responses to questions regarding rape myths and feminism would not contaminate responses to the remainder of the study.

Students who completed the attitude measures in pre-screening were eligible to register for a study titled “College Life: Responding to Stressful Events.” After registering, participants were emailed a link to the online study. When participants accessed the site they were instructed to read and respond to a scenario describing a college student’s negative experience. The scenario was ostensibly taken from a prior study in which students were asked to write about a stressful event that they experienced while at college. Participants were informed that a female student at the University of Connecticut wrote the following:

I went to an apartment party with a few of my girl friends. I was driving that night, so I wasn’t drinking. About an hour after I got to the party, I ran into this guy from class. We were talking about class and having a good time until I had to use the bathroom. The guy walked me to the bathroom and kept me company in line. When I went into the bathroom, he followed me in. He started coming on to me... kissing and touching me... and then he forced himself on me. But I didn’t want to have sex with him.

After reading the scenario, participants responded to a series of questions assessing their attributions of fault, their emotional reactions, and their desire to engage in collective action and reported likelihood of helping the survivor. These variables were measured after participants read the scenario because they represent state-level phenomena that are expected to be influenced by the situation described.

Measures

Attitudes Toward Feminism

Participants completed the Attitudes toward Feminism and the Women’s Movement Scale (FWM; Fassinger 1994) during pre-screening. The FWM Scale consists of 10 items designed to measure participants’ attitudes toward feminism, including “Feminist principles should be adopted everywhere.” Participants were asked to indicate the extent to which they agreed with these items on a Likert scale ranging from strongly disagree (1) to strongly agree (5). An average score of the items was created from the scale (α = .86).

Rape Myth Acceptance Attitudes

Participants also completed items from the Illinois Rape Myth Acceptance Scale (IRMA; Payne et al. 1999) during pre-screening. Experimenters were allowed to include only a limited number of items in pre-screening, therefore 10 out of 20 items were selected from the IRMA short form for inclusion. Items were chosen based on representativeness and potential for internal reliability. To achieve representativeness, one to two items were chosen from each of the seven subscales, which include: She asked for it (SA), It wasn’t really rape (NR), He didn’t mean to (MT), She wanted it (WI), She lied (LI), Rape is a trivial event (TE), and Rape is a deviant event (DE). To achieve potential for internal reliability, items that correlated highly with the entire scale in the original scale development study (Payne et al. 1999) were chosen for the current work. The following items were ultimately chosen based on these inclusion criteria: SA-3, WI-5, TE-5, WI-1, NR-1, LI-2, DE-3, LI-1, MT-1, and SA-1 (items can be found in Table 3 of Payne et al. 1999). An example item is “If a woman doesn’t physically fight back, you can’t really say that it was rape” (NR-1). Participants indicated the extent to which they agreed with the items on a Likert scale ranging from not at all agree (1) to very much agree (7). An average score of the items was created from the scale (α = .86).

Attributions of Fault

After reading about the scenario, participants were asked to indicate the extent to which they thought that society, the man involved in the scenario, and the woman involved in the scenario were each at fault for the incident. These items were adapted from a prior study examining predictors of collective action in response to violent crime (van Zomeren and Lodewijkx 2005). Participants were asked to rate how much they agreed that: “Society is responsible for the incident” and “Society can be blamed for the incident” on a Likert scale ranging from strongly disagree (1) to strongly agree (7). An average score of the two items was created (α = .91). These items were also employed to measure the extent to which participants attributed fault to the man and the woman by replacing “society” with “the man” (α = .85) and “the woman” (α = .78).

Emotional Reaction

Participants’ feelings of anger, fear, and pity were assessed. Emotional reactions items were also adapted from van Zomeren and Lodewijkx (2005)’s study, in which they were measured by asking participants to indicate the extent to which they felt several emotions when reading about an incident. In their study, anger items included “angry,” “furious,” and “powerless.” Pity items included “pity,” and “sorry.” Therefore, in the current study participants were asked to indicate the extent to which they felt these five emotions when reading about the incident described in the procedure on a Likert scale ranging from not at all (1) to extremely (7). To capture the extent to which participants felt fear, “fear” was also included. An exploratory factor analysis using principal axis factoring as the extraction method and oblimin as the rotation method was then conducted on the six emotion words. The factor analysis resulted in three factors that accounted for 71.4% of the variance. The first factor was labeled anger, and included “angry, and “furious.” The second factor was labeled fear, and included “fear,” and “powerless.” The third factor was labeled pity, and included “pity” and “sorry.” Average scores were created representing each of these factors (anger: α = .93; fear: α = .73; pity: α = .78).

Desire to Engage in Anti-rape Collective Action

Participants were asked to indicate the extent to which they would like to participate in and raise awareness about anti-rape efforts on their campus on a Likert scale ranging from not at all (1) to extremely (7). Items included: “March in Take Back the Night, a rally to protest rape and other forms of sexual violence,” “Forward an email advertising Take Back the Night,” “Join a group on campus that works to raise awareness about violence towards women,” and “Post a flyer in your residence hall advertising a group that raises awareness about sexual violence.” An average score of these four items was created (α = .88).

Reported Likelihood of Helping the Survivor

Participants were asked to imagine that the woman in the scenario was a friend or family member and indicated the likelihood that they would help the woman on a Likert scale ranging from not at all likely (1) to very likely (7). Items included: “I would do something to help the woman,” “I would offer to talk to her about her experience,” and “I would check in on her to see how she’s doing.” An average score of these three items was created (α = .88).

Results

Hypothesis 1: Gender Differences in Predictor Variables

We evaluated our first set of hypotheses by examining the means and standard deviations of each of the variables included in the study for women and men, and comparing them in a MANOVA analysis (see Table 1). We hypothesized that women would hold stronger feminist attitudes and weaker rape myth acceptance attitudes than men. This was supported: Women scored higher than men on attitudes towards feminism and lower than men on rape myth acceptance attitudes. We further hypothesized that women would attribute more fault to society, more fault to the male perpetrator, and less fault to the female survivor than men. This was partially supported: Women scored higher than men on attributions of fault to society. Women also scored lower than men on attributions of fault to the female survivor, however this gender differences were not statistically significant. Finally, we hypothesized that women would experience greater fear, anger, and pity than men. This was also partially supported: Women scored higher than men on fear and anger. Women also scored slightly higher than men on pity, however this gender difference was not statistically significant. Finally, these analyses revealed that women scored higher than men on desire to engage in anti-rape collective action, however there were no gender differences in reported likelihood of helping the survivor.

Table 1 Descriptive statistics and results of MANOVA comparing women and men
Full size table

We next explored the correlations between the variables for women and men (see Table 2). Among women, desire to engage in anti-rape collective action was correlated with the following predictor variables: attitudes toward feminism, rape myth acceptance, society at fault, man at fault, anger, fear, and pity. Among men, desire to engage in anti-rape collective action was correlated with the following predictor variables: attitudes toward feminism, rape myth acceptance, woman at fault, anger, and fear. Variables that correlated with desire to engage in anti-rape collective action among female and/or male participants were included in the first regression analysis. Among women, reported likelihood of helping the woman was correlated with the following predictor variables: rape myth acceptance, man at fault, anger, and pity. Among men, reported likelihood of helping the woman was correlated with the following predictor variables: attitudes toward feminism, man at fault, and anger. Variables that correlated with reported likelihood of helping among female and/or male participants were included in the second regression analysis.

Table 2 Correlations between variables: women are above the diagonal, men are below
Full size table

Hypothesis 2: Desire to Engage in Anti-rape Collective Action

Preliminary analyses indicated that the variables representing desire to engage in anti-rape collective action met the assumptions required for linear regression analyses: the dependent variable was normally distributed and the predictor variables lacked multicollinearity (all tolerances >.80, all VIF <2.00; Keith 2006). Therefore, a sequential linear regression analysis was conducted to examine the main effects of the predictors of desire to engage in anti-rape collective action as well as their interactions with gender (see Table 3). Sequential linear regression provides information about the percentage of variance in the dependent variable accounted for by the predictor variables at each step in the analysis. This is represented by the R 2 statistic. A perfect model that includes every relevant predictor variable would explain 100% (R 2 = 1.00) of the variance in the dependent variable, desire to engage in collective action. The R 2 statistic as well as the change in R 2 (Δ R 2) are provided in Table 3 for both blocks of the linear regression analysis of desire to engage in anti-rape collective action. Table 3 also includes both the unstandardized (B) and standardized (β) regression coefficients for the current analysis to better compare its results to the analysis of reported likelihood of helping the survivor.

Table 3 Sequential linear regression predicting anti-rape collective action
Full size table

At step 1, the main effects of gender and the predictor variables accounted for 36% of the variance in desire to engage in anti-rape collective action. We hypothesized that attitudes towards feminism positively predict desire to engage in anti-rape collective action. This was supported: Participants who were higher in attitudes towards feminism were more likely to desire to engage in anti-rape collective action. We also hypothesized that attributions of fault to society positively predict desire to engage in anti-rape collective action. This was also supported: Participants who attributed more fault to society for the rape were more likely to desire to engage in anti-rape collective action. Finally, we hypothesized that feelings of fear and anger positively predict desire to engage in anti-rape collective action. This was partially supported: Participants who felt more fear were more likely to desire to engage in anti-rape collective action, however feelings of anger were unrelated to desire to engage in anti-rape collective action. Additionally, the results demonstrate that female participants and participants low in rape myth acceptance attitudes were more likely to desire to engage in anti-rape collective action. At step 2, the interaction terms of gender and the predictor variables accounted for an additional 3% of the variance in desire to engage in anti-rape collective action. This was not a statistically significant increase in R 2, however, indicating that considering gender as a moderator did not help to explain desire to engage in anti-rape collective action. Further, none of the interaction terms are statistically significant, meaning that the slopes of the regression lines for female and male participants are parallel.

Hypothesis 3: Reported Likelihood of Helping the Survivor

Preliminary analyses indicated that the helping variable was negatively skewed (skewness = −1.30, SE = .18) because 40.8% of participants indicated that it was very likely that they would help the woman (i.e., their mean score on the helping variable was 7), but that the predictor variables lacked multicollinearity (all tolerances >.80, all VIFs <2.00; Keith 2006). Because skew represents a violation of one of the assumptions of linear regression, logistic regression was the appropriate analysis strategy for the helping variable (see Table 4). To conduct the hierarchical logistic regression, the helping variable was dichotomized to allow for a comparison between participants who indicated that it was very likely that they would help the woman (i.e., participants whose mean score on helping was 7) with the remaining participants who indicated at least some degree of hesitancy that they would help the woman (i.e., participants whose mean score on helping was less than 7). Similar to sequential linear regression, hierarchical logistic regression provides information about how well the predictor variables are able to correctly classify participants’ scores on the dependent variable. A perfect model that includes every relevant predictor variable would correctly classify 100% of the participants into the two categories represented by the dependent variable, reported likelihood of helping the woman. For the current model, the proportional by chance classification rate is .517. Therefore, the model would correctly classify 51.7% of the participants by chance alone.

Table 4 Hierarchical logistic regression predicting helping
Full size table

At block 1, 68.5% of participants were correctly classified by the main effects of the gender and predictor variables. We hypothesized that rape myth attitudes negatively predict reported likelihood of helping a survivor of rape. This was supported: Participants who were higher in rape myth acceptance attitudes were less likely to report that they would help the woman. We also hypothesized that attributions of fault to a female survivor of rape negatively predict and attributions of fault to a male perpetrator positively predict reported likelihood of helping a survivor of rape. This was partially supported: Participants who attributed fault to the man were more likely to report that they would help the woman, however attributions of fault to the woman were unrelated to reported likelihood of helping the woman. Finally, we hypothesized that feelings of anger and pity positively predict reported likelihood of helping. This was also partially supported: Participants who felt more anger were more likely to report that they would help the woman, however feelings of pity were unrelated to reported likelihood of helping. Additionally, female participants were more likely to report that they would help the woman. At block 2, an additional 4.5% of participants were correctly classified with the addition of the interaction terms between gender and the predictor variables. The interaction between attributions of fault to the man and gender was marginally statistically significant. Further analyses demonstrated that female participants who attributed more fault to the man were more likely to help the woman, B (SE) = 1.23 (.31), Exp(B) = 3.44, Wald = 16.02, p < .001. However, there was no relationship between male participants’ attributions of fault to the man and their reported likelihood of helping the woman, B (SE) = .43 (.32), Exp(B) = 1.54, Wald = 1.79, p = .18. The entire model correctly classified 73.0% of participants on the helping dependent variable. This model performed 41.2% better than chance alone (i.e., 51.7%).

Discussion

In the present work, we focused on identifying the factors that are related to American college students’ desire to engage in anti-rape collective action and reported likelihood of helping a survivor of rape. More specifically, we explored how attitudes, attributions of fault, and emotions are related to these two forms of responses to rape among women and men. Although past research has examined these three factors individually, we examined them simultaneously to learn which ones act as predictors of responses to rape when controlling for the others. The results of this work suggest that these factors are differentially related to desire to engage in anti-rape collective action and reported likelihood of helping survivors of rape. Gender, attitudes toward feminism, rape myth acceptance attitudes, attributions of fault to society, and feelings of fear emerged as predictors of desire to engage in anti-rape collective action. Gender did not moderate these effects, suggesting that the relationships between these predictors and desire to engage in anti-rape collective action were similar for female and male participants. In contrast, gender, rape myth acceptance attitudes, attributions of fault to the man, and feelings of anger emerged as predictors of reported likelihood of helping the survivor. Gender moderated the effect of attributions of fault to the man, indicating that women’s attributions of fault to the man were positively related to their reported likelihood of helping the survivor but men’s attributions of fault to the man were unrelated to their reported likelihood of helping the survivor. In sum, the current work indicates that attitudes, attributions of fault, and emotions contribute independently to college students’ responses to rape and that the predictors of these responses are different.

Insights into Intended Responses to Rape

The results of the current study highlight the important role that attributions of fault play in American college students’ responses to rape. Although past research on attributions of fault within rape scenarios has largely focused on the conditions under which people blame the female survivor (Barnett et al. 1992; Cowan and Quinton 1997; Gillen and Muncer 1995; Grubb and Harrower 2008; Kanekar and Seksaria 1993), the current work draws attention to attributions of fault to society and the male perpetrator. Attributions of fault to society were related to desire to engage in anti-rape collective action. Students who attribute fault to society for rape may desire to engage in anti-rape collective action in hopes of bringing attention to and ultimately dismantling societal forces contributing to rape. In contrast, attributions of fault to male perpetrators were related to reported likelihood of helping survivors. Therefore, students who make this individually-based attribution of fault may want to engage in an action to restore the well-being of the individual survivor. Interestingly, this finding was moderated by gender: attributions of fault to the male perpetrator were related to reported likelihood of helping among female participants but not male participants. Although past research has demonstrated that female college students attribute more fault to male perpetrators than male college students (Krulewitz and Nash 1979), it is unclear why these attributions would be related to reported likelihood of helping the survivor among female students but not male students. This finding might therefore reveal an especially interesting avenue for future research.

The current study further highlights the importance of emotional reactions in American college students’ responses to rape. Past research has demonstrated a connection between feelings of anger with engagement in non-feminist collective action among U.S. adults (Smith et al. 2008) and Dutch college students (van Zomeren et al. 2004; van Zomeren and Lodewijkx 2005), as well as feminist collective action among Australian adult women (Hercus 1999). Anger, however, was not a predictor of U.S. college students’ intentions to engage in anti-rape collective action in the current study. Instead, fear was a predictor of intentions to engage in anti-rape collective action. This suggests that the emotional process involved in responses to anti-rape collective action may differ from other forms of collective action. Anger was instead related to reported likelihood of helping the survivor. This finding is consistent with past research that has demonstrated a relationship between feelings of anger and helping (van Zomeren and Lodewijkx 2005). Finally, although participants felt pity towards the survivor, feelings of pity were unrelated to responses to rape after controlling for the remaining predictors. Future work might therefore further explore the role of fear in anti-rape collective action and anger in helping survivors of rape.

Limitations of the Current Work

There are limitations to the current study that are important to consider when interpreting its results. For example, the rape scenario described in the study may not be typical of rape that occurs on college campuses. In particular, the woman described in the scenario was not drinking at the party, which may be unusual for both a college party and a rape scenario. This description likely influenced participants’ perceptions of the woman, making them less likely to see her as at fault for the rape. The woman was also not described as engaging in other behaviors that often lead to victim blame. It is possible that participants’ attributions of fault would have been different given the constraints of a different scenario and therefore may have been related to responses to rape.

Further, the current study examined participants’ desire to engage in anti-rape collective action and reported likelihood of helping the survivor described in the scenario rather than actual engagement in anti-rape collective action and helping survivors of rape. Therefore, constructs closer to behavioral intentions rather than actual behaviors were measured. It is possible that actual behavior in response to rape might deviate from reported intended behavior. There is reason to believe, however, that these behavioral intentions are an appropriate proxy for actual behavior. The theories of reasoned action (Fishbein and Ajzen 1975) and planned behavior (Ajzen 1985) provide a theoretical rationale supporting the relationship between behavioral intentions and actual behavior. This theoretical rationale is supported by empirical work demonstrating that behavioral intentions lead to actual behavior (Ajzen and Fishbein 1973; Hagger and Chatzisarantis 2009; Schifter and Ajzen 1985). It is therefore plausible that the desire to engage in anti-rape collective action and reported likelihood of helping the woman captured in the current study would be related to actual completion of these behaviors.

An additional limitation of the current work is the extent to which its findings are generalizable. The goal of the work was to identify the factors that predict American college students’ behavioral responses to rape. The sample, however, was comprised of mostly white college students attending a large public university in New England and therefore may not be representative of all American college students, particularly racial minorities. Future work might explore whether the factors that were identified in the current study are generalizable to other American college students.

Conclusions

Given that rape is a significant problem on college campuses in the U.S., it is important to better understand American college students’ responses to rape. These responses may be directed at societal forces that perpetuate violence towards women. They may play an important role in raising awareness about sexual assault and rape, and ultimately contribute to a lessening of sexual violence. Additionally, these responses may be directed at individual survivors of rape. College students’ responses to survivors could have a significant impact on survivors’ psychological and physical well-being. The current study suggests that American college students’ desire to engage in anti-rape collective action and reported likelihood of helping a survivor of rape are related to their attitudes, attributions of fault, and emotions but that the specific attitudes, attributions of fault, and emotions related to each response are different. The current study further suggests that the processes whereby these factors relate to intended responses to rape are similar for female and male college students. A stronger, more nuanced understanding of the predictors of college students’ desire to engage in anti-rape collective action and likelihood of helping survivors of rape will contribute to efforts to stop rape on college campuses in the long-term and improve the outcomes of students who have survived rape in the short-term.